[Bf-committers] Re: Minnaert diffuse shader

[Alfredo de Greef]

GSR wrote:
> ....
> And talking about most sources:
>
http://www.cs.princeton.edu/~smr/cs348c-97/surveypaper.html
> And nvidia "flipped" method:
>
http://cvs1.nvidia.com/DEMOS/OpenGL/src/cg_lightmodel/Surface%20Reflection%20Models.doc
> ....

The Minnaert model was one of the things I tried when
I first got into creating texture plugins for Blender.
Of course that wasn't possible at the time (because of
the missing required info, though I found a way later
to do it anyway).
In any case, the model as implemented by lordloki is
the one from the nvidia doc as he said himself.
I have tried both in the past, and I still don't know
which one is supposed to be correct. The 'official'
model can also return values much higher than 1 (at
grazing angles), making it look more like a specular
model sometimes.
The nvidia model behaves better, and (to me) looks
most like velvet when k ('darkening limb') is 0.
This is also basically equivalent to using a ramp from
white to black normal mapped and in multiply mode, or
a z-only normal mapped sphere blend texture similar to
how fresnel effects used to be simulated in Blender.
In my own programs sofar I have only used the official
model, not the nvidia one. I suppose the best option
is to implement both and let users decide.

Alfredo



	
		
__________________________________ 
Do you Yahoo!? 
Yahoo! Mail - You care about security. So do we. 
http://promotions.yahoo.com/new_mail